Actually that is not how statistics work. Confidence level does not tell you whether your claim is in the “ball park” of the real value. It is rather a way to asses the validity of your claim. In most statistical studies (even in the social sciences), confidence levels at 95% is the minimum needed to assume the validity of your hypothesis. That is two standard deviations from the mean. A 68% confidence level would generally be considered insignificant, the hypothesis was not supported.
If the hypothesis that higher interest rates lowers rent inflation is only supported by a 68% confidence level, then the standard practice would be to reject the hypothesis.
To sum up, we do not know with high confidence whether raising interest rates will lower rent inflation.